Patients with type 1 diabetes mellitus (T1DM) will experience dramatic improvements in their quality of life and life expectancy when the goal of engineering an artificial pancreas is finally realized. A critical component of such a system is to be able to measure changes in blood glucose rapidly, accurately, and continuously. Implantable continuous glucose monitors are an improvement over the fingerstick measurements that have been the norm for decades, but their performance is hampered by fundamental limitations of the interstitial (subcutaneous) space where they are currently implanted: 1) there is substantial lag time between changes in blood glucose and changes in interstitial glucose, 2) there is substantial variability in the lag time because perfusion of the interstitial space is variable, 3) there are steady-state inaccuracies due normal fluctuations in interstitial oxygen tension, and 4) due to tissue encapsulation, interstitial sensors need to be replaced frequently. These limitations prevent interstitial sensors from being used to achieve tight glycemic control around routine activities that involve rapid changes in blood glucose, such as eating and exercise. Achieving tight glycemic control is the long-term goal of our research program, as it is critical to preventing the devastating long-term sequelae in patients with T1DM. The objective of this application is to determine the extent to which using the peritoneal space for glucose sensing (instead of the interstitial space) will move us toward this goal. Our central hypothesis is that the fluid in the peritoneal space tracks blood glucose changes with less lag time and less lag-time variability than the interstitial space, because the blood flow to this central, protected space is copious and robust to changes in temperature and cardiac output, and because the glucose kinetics of the peritoneal space are known to be fast. We additionally hypothesize that peritoneal sensors will exhibit less intersensor variability and oxygen- tension-related inaccuracies than the same sensors placed interstitially. Our hypothesis is supported by the physiology literature and by our pilot studies using a glycemic challenge, which show a faster glucose response for the peritoneal space vs. interstitial sensors. We will test these hypotheses by comparing continuous glucose readings from sensors implanted in the peritoneal vs. interstitial spaces while exposing experimental animals (pigs) to intravenous glucose tolerance tests. These tests will be done under baseline conditions, then again during changes in body temperature and blood pressure, to test robustness. We will measure lag times and lag-time variability in all cases. Additionally, we will test our proprietary technology for preventing tissue encapsulation, which is one of the principle challenges facing efforts to implant sensors chronically.